Start a new topic

REQ: Better automatic tag placement

Source segment (the tags correspond to CT tags here):

{3}<x2></x2><x3></x3>Rue de la Tour 2<x4></x4><x5></x5>{4}<x6></x6><x7></x7>Phone<x8></x8><x9></x9>{5}<x10></x10><x11></x11>+41 55 666 66 66


Target segment:

{3}<x2></x2><x3></x3>Rue de la Tour 2<x4></x4><x5></x5>{4}<x6></x6><x7></x7>Telefon<x8></x8><x9></x9>{5}<x10></x10><x11></x11>+41 55 666 66 66


So far, so nice. Then there is

Target segment:

{16}<x2></x2><x3></x3>Rue de la Tour 2<x4></x4><x5></x5>{17}<x6></x6><x7></x7>Phone<x8></x8><x9></x9>{18}<x10></x10><x11></x11>+41 55 666 66 66


The numbers in bold letters are the only change (indeed, they are not bold in the source, only to make it clearer here)


But the target is now propagated as follows (observe the tag positions, including again a tag in a word):

{16<x2></x2><x3></x3>}Rue de la Tour <x4></x4><x5></x5>2{1<x6></x6><x7></x7>7}Telef<x8></x8><x9></x9>on{<x10></x10><x11></x11>18}+41 55 666 66 66


Let's call it a tag soup. Of course, I could turn off auto propagation and tag guessing (like in a garage "please only use the first gear"). However, this is hilarious. 


Would it at least be possible to avoid tag positioning inmidst a word, and the better positioning with numbers could perhaps also be extended to parentheses.





There is no so much tag soup here. You made it yourself for whatever reason. Please switch on adjacent tags merging in Action > Tags menu. First, it will facilitate the program to auto-propagate the tagged numbers. Second, it will make your point clearer.


To use your car analogy,  "please slow down on bumpy roads".

I should have said that. "Adjacent tags merging" was on from the beginning, this citation above is simply what is happening when you copy and paste it from CT. See below the result again:


image


To use my car analogy: "no, this is what happens on very smooth and straight highways".



Just to comment this figure:

The automatic tag positioning seems to base merely on the character position in the first segment and does not take account of parentheses or numbers (or the non-interruption of words). A bit more sophistication would be nice.


1 person likes this
Login to post a comment